Relevance of plant stress monitoring
As the human population grows, the efficiency of agriculture becomes a crucial issue. Crop production, even though it is noticeably different from what it was decades ago, still can and will be improved given a pressing need for food security and active research in agriculture.
A simple approach of sowing larger areas with crops not only is an expensive endeavor but also comes with a list of undesirable consequences, such as natural habitat destruction, soil degradation, and simply limited area that can be utilized for agricultural purposes.
A better way to deal with this problem is to minimize crop losses on the existing fields. There are numerous harsh conditions to which plants can be subjected during their growth which usually results in plants’ premature death or growth anomalies.
The first method used to fight crop loss due to these stresses is genetic modification of the crop (known as GMO — Genetically Modified Organisms) in a way that makes a plant stress-resistant. However, GMO technology is not a panacea to all agricultural challenges and has its challenges and drawbacks.
All concerns mentioned above leave the best option of all: early detection and fighting a particular stressor. Monitoring can also be done on different scales and levels. The most scalable and cost-effective ones are the use of satellite imagery and the use of UAVs (Unmanned Aerial Vehicles). The use of in-field sensors might get costly and troublesome, especially on a large scale, however.
Different stresses and their sources
Plant stresses are usually divided into two groups: biotic (stresses resulting from exposure and interactions with other living organisms) and abiotic (stresses that arise from environmental conditions and characteristics).
Biotic stresses can involve any exposure to viruses, fungi, or bacteria which can either cause severe disease or prevent natural plant growth due to exposure to parasites. Competing plants might also pose issues in plant development, depriving them of nutrients and water.
Among abiotic stressors, plants can be subjected to temperature extremes, droughts. Water availability is undeniably the most critical factor as it is used by plants for nutrient transportation and serves as a precursor for energy storage in the form of glucose. The temperature extremes can damage a plant’s protective cover negatively, impacting a normal process of photosynthesis.
Chemical characteristics and composition are also crucial for healthy vegetation to develop. Moreover, one soil might be more suitable for one crop but not the other purely because of such characteristics.
The most accurate assessment: laboratory tests. Pros and cons.
Laboratory tests are usually used for high-quality checks if a plant is subjected to one stressor or another. Determination of chemicals contained in the soil or plant can be known with high precision in laboratory settings. If there is any biotic stressor, it can be determined what particular species of insect, fungus, or bacteria is the main danger for the crop.
These are certainly pros. What about cons?
Laboratory tests are disruptive, meaning one has to physically remove a plant to test it, remove a portion of the soil, or anything subjected to the test for that matter. Another issue with “in situ” tests is the inability to scale; the procedures of tests are usually expensive and time-consuming, while results of the test generally apply to a small population or area.
There is no doubt that it is a good idea to obtain a sample from the soil and do an accurate assessment of all needed attributes once or twice per season. But for intermediate monitoring, it is not a suitable choice.
Methods of monitoring and assessment with remote sensing
Visible and infrared light interacts differently with matter depending on its chemical content and structure. To create an even more informative picture, different specters can be combined into reflectance indices.
The most important one being the Normalized Difference Vegetation Index (1). A healthy plant during warm seasons of blooming reflects a large portion of near-infrared radiation, pushing such ratio closer to 1, write the opposite is true when the plant is affected by stressors. Normalized Difference Moisture Index (2) is a similar index which helps to estimate crop water stress level. Normalized Difference Red Edge (3) usually is correlated with nitrogen content in plants’ stomata. There are a lot of other spectral indices to go through all of them, each of which might provide some aspect of plant stress regarding one or several problems.
An example of the aforementioned indices, along with the corresponding RGB image is shown in figure 1.
NDVI = (NIR — RED)/(NIR + RED) (1)
NDMI = (NIR — SWIR)/(NIR + SWIR) (2)
NDRE = (NIR — RE)/(NIR + RE) (3)
NIR — reflection in the near-infrared spectrum (785–899 nm),
RED — reflection in the red range spectrum (650–680 nm),
SWIR — reflection in the short-wave infrared spectrum (1565–1655 nm),
RE — reflection in the near-infrared spectrum (698–713 nm)
In much the same way as visual and infrared imaging, hyperspectral imaging allows obtaining detailed information regarding reflected waves. It deserves a separate category of imaging due to its very narrow range of wavelengths binning. For comparison, Sentinel-2 mission satellites have 13 bands ranging from 458 nm wavelength up to 2280 nm, while Earth Observer-1 (EO-1) Hyperion Mission provides images with wavelengths in the range from 357 to 2576 nm binned in 220 separate channels.
Not only satellites can provide such data, UAVs as well were widely equipped with sensors of different kinds and used for hyperspectral imaging. Such data, without doubt, might be valuable for agricultural applications.
Despite being a great source of information about waves’ interaction with molecules of plants’ leaves and soil surface, even infrared imaging cannot estimate with high accuracy one critical aspect in plant stress assessment, namely, temperature. There has been a great deal of research effort spent and proposed several spectral indices that take into account temperature data besides the usual visible or infrared radiation data, such as Crop Water-Stress Index (CWSI), which usually performs better in water content estimations.
Just like molecules’ interaction with waves of infrared wavelengths has a unique signature, waves or ultraviolet range help to conclude chemical and physical properties of plants and soils. Fluorescent images have been proven to be helpful in early crop disease detection, heavy metals content, and soil salinity estimation. Ironically, a laboratory test for chlorophyll content estimation is a very similar procedure known as spectroscopy. Quite rarely, this value is derived with other tests, such as chemical reactions to dioxane.
In the second part of the article, we will focus on stresses forecasting, data resources, their quality, and other issues. Stay tuned!